Six metrics that predict a good hire — with thresholds, red flags, and the cost of getting it wrong. Use them on any supplier. Including us.
From "IT Staffing bez straconych szans" — a workshop with technology & procurement leaders, early 2026 · Published August 2026 · Last updated August 2026
We opened the workshop with one question: what is your biggest pain working with external IT suppliers? These are the answers, unedited, grouped the way the room grouped them.
"We get CVs that don't match the requirements at all. The candidates don't know the technology from the job ad."
"The recruiter doesn't understand the difference between a senior and a mid-level in our context. The CVs are correct on paper, but not in the interview."
"I get the impression the agency sends us everyone in their database, hoping we'll do the filtering ourselves."
"I don't know whether our process is fast or slow relative to the market. There's no point of reference."
"By the time we decide on a candidate, they've already signed somewhere else. It keeps happening."
"I have no information about what happens after I send the requirements. I either get CVs or I don't — no status."
"The vendor is very active right up to signing. Then they disappear. No check-ins, no follow-up."
"The engineer left after two months. The agency said 'we're sorry' and offered a new one. No analysis of what went wrong."
"The only communication is the invoice."
None of this is a shortage of good developers. It's a shortage of measurement. The six KPIs below fix that.
Measure every supplier on all six. Five measure the vendor. The sixth — deliberately — measures you.
Time from briefing to the first quality CV. Shows whether the vendor has a ready candidate pool or starts sourcing from zero on your order.
If the first CV shows up after a week, the vendor wasn't ready for your order. Business days.
The percentage of submitted CVs you actually invite to interview. The cleanest measure of pre-screening quality.
Below 40% means mass-sending without selection. Your time has become the vendor's free filter.
Time from briefing to a signed contract with the candidate. Measure it separately on the vendor's side and your own.
Long time-to-hire is more often caused by slow client feedback than by a lack of candidates — check Feedback Loop Time before blaming the vendor.
The percentage of offers candidates accept. A low rate means the vendor isn't managing candidate expectations on terms and conditions.
If candidates learn the contract terms only at offer stage, offers get declined. That's a vendor error, not bad luck.
The percentage of hires still in place after 90 days. The single most important measure of long-term match quality.
High churn in the first three months signals a matching error — and responsibility can sit on both sides of the process.
Time from CV submission to your feedback reaching the vendor. The one KPI on this list that measures the client, not the supplier.
Slow feedback is the single biggest reason the best candidates are lost. A senior developer typically has three or four live processes running at once.
The workshop's sharpest finding: how you buy predicts the outcome more reliably than who you buy from. A real before/after from one hiring process:
This is why the hourly rate is the last thing to optimise in a tender, not the first. Companies that squeeze the rate pay a multiple of the difference in hidden costs: project delays, engineer churn, and leadership time spent screening CVs.
From our own delivery records — a senior cybersecurity (WAF) search, February 2024, full funnel:
What the funnel taught: the offer, not the market, was the constraint — the rate cap sat below senior Warsaw expectations and the office requirement halved the pool. We told the client at the brief, not in the post-mortem. The full breakdown lives on the Cybersecurity page →
"We have every competence." No specialisation usually means sourcing gets pushed outside the moment you sign. Specialisation is credibility.
First CV arriving after more than five working days means there is no ready pool — the vendor is sourcing from zero, for you, on your clock.
Nobody accountable for introducing the candidate. A problem with the engineer after start is met with silence on the vendor's side.
The same profiles arriving from several suppliers at once. No exclusivity means no real engagement in your process.
We built this framework, so it's only fair we're measured by it. These are the commitments we sign against — and invite you to score:
Scorecard weights: quality 30% · speed 25% · retention 20% · communication 15% · transparency 10%
Between three and five. If you are receiving forty CVs for one role, the supplier is not pre-screening — they are using your time as a free filter. The benchmark to watch is CV-to-Interview Rate: above 60% is good, below 40% means mass-sending without selection.
Under four weeks is good; four to eight weeks is average; more than eight weeks signals a problem. Measure it separately on the supplier side and your own — long time-to-hire is more often caused by slow client feedback than by a shortage of candidates.
Across five weighted dimensions: candidate quality (30%), speed of response (25%), retention (20%), communication (15%) and transparency (10%). Score each supplier A–D per criterion, multiply by the weights, and review quarterly. Anything below 60% overall needs a remediation plan or a replacement. The full scorecard is available as a free, vendor-neutral template.
Workshop-derived estimate: 15–25% of the annual salary in total, once you count 8–12 weeks of lost process, 40–80 hours of leadership time, and 3–6 months of project delay while the role sits open. It is why the hourly rate is the last thing to optimise, not the first.
Put a slot in a calendar, or send the brief and we will come back to you. Either way you are talking to a partner, not a queue.
Thirty minutes, no deck. Whoever's calendar suits you — they take the call themselves.
Roles, platform, headcount, timeline. A link to a brief is even better.